Ординатура / Офтальмология / Английские материалы / Eye, Retina, and Visual System of the Mouse_Chalupa, Williams_2008

is indirect. Three likely target genes for Pou4f2 have been identified to date. One is Pou4f1, which is closely related to Pou4f2 and is expressed in RGCs one day later than Pou4f 2. Turner and co-workers have reported the presence of a regulatory region upstream of Pou4f1 containing Pou4fbinding sites that can be activated by either Pou4f1 or Pou4f2 (Trieu et al., 1999). It is possible that Pou4f2 initially activates Pou4f1, and its expression is then maintained by both Pou4f1 and Pou4f2.

Our laboratory has recent results using electrophoretic mobility shift assays, transient transactivation of reporter genes in tissue culture cells, and chromatin immunoprecipitation analysis with embryonic retinas suggesting that Pou4f2 binds to and activates transcription from Pou4f-binding sites within regulatory regions associated with the genes encoding Tbr2/Eomes, a T-box transcription factor, and sonic hedgehog (Shh), a secreted signaling molecule (Mu et al., 2004; Mao et al., 2008). Recently, techniques have become available to perform genome-wide chromatin immunoprecipitation analysis for global mapping of transcription factor binding sites and target gene networks (Ji et al., 2006; Zeller et al., 2006). These techniques are ideally suited for genomewide identification of Math5and Pou4f2-binding sites, as well as binding sites for other transcription factors in the RGC gene regulatory network.

Regulation of Retinal Ganglion Cell Axon Outgrowth and Pathfinding Retinas from Pou4f2 null mice contain residual RGCs that survive into adulthood and extend neurites (Gan et al., 1996, 1999; Wang et al., 2000). However, these neurites do not grow normally and have defects in finding their appropriate targets in the brain (Erkman et al., 2000; Wang et al., 2000, 2001). Several genes whose products are involved in axon outgrowth and pathfinding are strongly downregulated in the absence of Pou4f2. These include axon guidance factors Gap43 and neuropilin1, and several components of the axon cytoskeleton, including persyn/ gamma-synuculein, neurofilament light chain, neurofilament middle chain, and tau (Mu et al., 2004). These results suggest that Pou4f2 directs a hierarchical program linking signaling events to cytoskeletal changes required for axon outgrowth and pathfinding (Erkman et al., 2000). Recent results from our laboratory indicate that Tbr2/Eomes, which appears to be a direct target of Pou4f2, functions in this hierarchical program. RGCs and optic nerves from Tbr2/Eomes knockout mice have defects similar to those of Pou4f2 knockout mice, although somewhat milder (Mao et al., 2008). This suggests that Pou4f2 works partly through Trb2/Eomes to regulate axon outgrowth and pathfinding.

Other aspects of RGC axon outgrowth, pathfinding and retinotectal topographic mapping may depend on regulatory processes not controlled by Pou4f2. Both canonical and noncanonical Wnt signaling have been shown to be

involved in the growth and retinotectal projections of RGC axons (Ouchi et al., 2005; Rodriguez et al., 2005; Liu et al., 2006; Schmitt et al., 2006). In these examples, Wnt signaling components are expressed in RGCs, and therefore their expression must be regulated by RGC transcription factors.

Control of Retinal Progenitor Cell Proliferation by Retinal Ganglion Cell–Secreted Signaling Molecules

In the past few years, studies from a number of laboratories have revealed a previously unsuspected function for RGCs in regulating the proliferation rate of RPCs. These studies have shown that in addition to their well-known physiological functions in receiving, processing, and transmitting visual information to the brain, RGCs are also a source of several secreted signaling molecules that modify the extrinsic environment of the developing retina. One of the most important signaling molecules in early retinogenesis is Shh. Wallace and co-workers have demonstrated that RGCderived Shh controls RPC proliferation and the timing of RGC development in the developing mouse retina (Dakubo and Wallace, 2004; Wang et al., 2005). Shh signaling upregulates the gene encoding CyclinD1, which is known to play an important role in RPC proliferation (Sicinski et al., 1995); in retinas where the Shh gene is conditionally knocked out, CyclinD1 is strongly downregulated (Wang et al., 2005). In zebrafish retinas, Shh produced by newly differentiating RGCs is thought to be required for the central-peripheral propagation wave along which RGCs and other retinal cell types differentiate (Neumann and Nüesslein-Volhard, 2000), but this has not been supported by other studies (Kay et al., 2005; Masai et al., 2005). However, in contrast to the role played by Shh in stimulating RPC proliferation in the mouse retina, Shh appears to negatively regulate the cell cycle and to promote cell cycle exit in the zebrafish retina (Masai et al., 2005; Neumann, 2005). Although the reasons for the discrepancy between zebrafish and mouse are not clear, a more recent report with Xenopus retinas indicates that Shh may modulate the cycling rate of RPCs, thereby promoting cell cycle exit (Locker et al., 2006).

Gene expression profiling with Pou4f 2 null and Math5 null retinas has identified three secreted signaling molecules whose genes are expressed exclusively in RGCs during retinogenesis (Mu et al., 2004; Mu et al., 2005b). Shh and Gdf8/Myostatin, a close relative of Gdf11, were strongly downregulated in Pou4f 2 null and Math5 null retinas, whereas Vegf was downregulated only in Math5 null retinas. Notably, the absence of Pou4f2 does not alter Vegf expression, indicating that Vegf is regulated by a non-Pou4f2 branch of the RGC gene regulatory network (see figure 26.2). As discussed earlier in the chapter, it is likely that Shh is a direct target gene of Pou4f2 (Mu et al., 2004; Fu et al., 2008). RGC-derived Shh is thought to function by binding to the Patched receptor on

the surface of RPCs and initiating a signal transduction pathway mediated through the Gli family of transcription factors. Gli1 and Gli3 are expressed in RPCs (Dakubo and Wallace, 2004; Mu et al., 2004), and Gli1 expression is under the control of the Shh signaling pathway, which reinforces the Shh signal by positive feedback. Gli1 expression in RPCs is dramatically downregulated in the absence of Shh or Pou4f2 (Dakubo and Wallace, 2004; Mu et al., 2004; Wang et al., 2005). This suggests a model in which Pou4f2 activates the expression of Shh in RGCs thereby leading to the secretion of Shh into the extracellular environment. Secreted Shh then binds to Patched on the surface of RPCs, and Gli1 expression is induced.

In the chick retina, Vegf appears to stimulate RPCs by activating the MEK-ERK pathway as well as by elevating Hes1 expression (Hashimoto et al., 2006). Vegf ligand secreted by RGCs works through the Flk1 receptor located on the surface of RPCs (Hashimoto et al., 2006). The results with Shh and Vegf suggest that RGCs act in the ganglion cell layer as a developmental signaling center to influence RPC proliferation rates. Other secreted signaling molecules, such as GDF8/myostatin and GDF11, may also contribute to signaling events that are initiated from newly differentiated RGCs.

In a recent investigation, our laboratory asked whether the ablation of RGCs from the developing retina would alter the extracellular environment. We were particularly interested in determining whether the loss of RGCs affected the ability of other retinal cell types to assume their fates. Models have been proposed in which RGCs and other retinal cell types provide feedback signals to the RPCs to alter the extrinsic environment and allow RPCs to advance to new competence states (reviewed in Cayouette et al., 2006). We ablated RGCs genetically by inserting a conditionally expressed gene encoding diphtheria toxin A (DTA) polypeptide into the Pou4f2 locus and activating the Dta gene early in retinogenesis using a Six3-Cre recombinase mouse line (Furuta et al., 2000; Mu et al., 2005b). RGCablated retinas produced all of the retinal cell types (with the exception of RGCs) in the correct locations and relative proportions. However, the retina was 30%–50% thinner as a result of attenuation in the rate of RPC proliferation (Mu et al., 2005b). Genes encoding cell cycle regulators CyclinD1 and Chx10 are downregulated in RGC-ablated retinas, while Math5 expression is significantly upregulated. Apparently, the continuous loss of RGCs alters the intrinsic program of RPCs such that they abnormally exit the cell cycle and express Math5 in a futile attempt to generate more RGCs. The results are consistent with previous findings that secreted signaling molecules produced in RGCs regulate the rate of RPC proliferation and control RGC number and timing of differentiation. The results of Mu et al. (2005b) demonstrate that RGCs are not required for RPC

competence change and suggest that intrinsic rather than extrinsic factors play the major role in specifying retinal cell types. However, RGCs do affect RPC proliferation, and this provides a positive feedback mechanism to ensure there is a sufficient supply of RPCs throughout retinogenesis (Mu et al., 2005b).

Future prospects

Developmental Regulation of Retinal Ganglion Cell

Subtypes Throughout this review, we have considered RGCs as a uniform cell type, which is an oversimplification. Numerous RGC subtypes have been identified by their morphological and physiological properties in the retinas of several vertebrate species. For instance, RGC axons have different retinotectal projections along the nasal-temporal and dorsal-ventral axes, indicating a functional specialization associated with axial patterning (McLaughlin et al., 2003). Unfortunately, the number of RGC subtypes that are present in the mouse retina is ill-defined, and a lack of agreed-upon criteria for identifying RGC subtypes makes it difficult to address the mechanisms by which they develop. In a recent study, Coombs et al. (2006) used 14 morphological measures to classify mouse RGCs parametrically into different clusters. Their analysis revealed 14 clusters with distinct morphological features. The ability to apply straightforward, reproducible criteria to identifying RGC subtypes should yield a valuable baseline for future studies on RGC development (Coombs et al., 2006).

Because of a lack of functional and molecular definition for most RGC subtypes, virtually nothing is known about the mechanisms that control their formation, although a few studies have described the development of morphological diversity during postnatal stages of retinogenesis (e.g., Diao et al., 2004; Mumm et al., 2006). One extensively investigated RGC subtype was initially identified by its expression of a gene encoding a novel photoreceptor, melanopsin, which was shown to be involved in nonvision light responses (Hattar et al., 2003). Melanopsin-expressing RGCs make up about 1% of mouse RGCs and function in circadian photoentrainment and pupil constriction (Hattar et al., 2003). How Melanopsin-expressing cells arise is not yet understood, but in Math5 null retinas, photic entrainment is lost and circadian rhythms are altered, suggesting that Math5 regulates the development of these cells (Wee et al., 2003).

In Vitro Generation of Retinal Ganglion Cells from

Embryonic Stem Cells An ultimate objective in research on retinal development is to establish practical methods for repairing and regenerating damaged retinas. A striking example of the progress being made in this area is the recent work of MacLaren et al. (2006), who transplanted

photoreceptor precursor cells from late ontogenetic stages into genetically damaged adult mouse retinas and achieved a remarkable degree of repair, as evidenced by the differentiation of the precursors into rod photoreceptor cells, the formation of synaptic connections, and improved visual function. This study emphasized the importance of selecting the appropriate progenitor cells for transplantation. One way to enhance the suitability of RPCs for transplantation is to manipulate them genetically so that they express combinations of transcription factors that confer the desired characteristics. Introducing different combinations of transcription factors associated with different retinal cell fates into naive progenitor cells might be a useful approach to generate precursor cells for transplantation. In a study using Xenopus, coexpression of various combinations of transcription factors transfected into developing retinas resulted in retinas in which cell fates were significantly altered according the predicted combinatorial coding (Wang and Harris, 2005). Given our current knowledge of the RGC gene regulatory network, it should now be possible to genetically alter uncommitted progenitor cells using key transcription factors such as Pax6, Math5, Pou4f2, and others to coax progenitors into assuming RGC fates with high efficiency.

A promising method for generating retinal cell types in vitro is through selective differentiation of embryonic stem (ES) cells. This has been achieved by supplying ES cells with the appropriate growth factor medium. The generation of retinal precursors and differentiated retinal neurons has been reported using mouse ES cells (Zhao et al., 2002; Ikeda et al., 2005), and efficient generation of retinal precursors that differentiated primarily into RGCs and amacrine cells was recently reported using human ES cells (Lamba et al., 2006). Procedures for introducing genes into ES cells are well established, thus providing a means to generate ES cells that express transcription factors positioned at key nodes in the RGC network. Within the foreseeable future, it should be possible to efficiently generate large fields of RGCs by differentiating ES cells in vitro with various genetic manipulations.

acknowledgments Work was supported by grants nos. EY11930 and EY10608 from the National Eye Institute, and by grants from the Ziegler Foundation and the Robert A. Welch Foundation, G-0010. The generation of genetically engineered mice, maintenance of mice, and DNA sequencing were supported in part by a National Cancer Institute Cancer Center support grant no. CA16672. We are grateful to Nadean Brown of the Children’s Hospital Research Foundation, the University of Cincinnati, and Xueyao Fu and Jang-Hyeon Cho of the M. D. Anderson Cancer Center for allowing us to cite their unpublished results.

REFERENCES

Amore, G., Yavrouian, R. G., Peterson, K. J., Ransick, A., McClay, D. R., and Davidson, E. H. (2003). Spdeadringer, a sea urchin embryo gene required separately in skeletogenic and oral ectoderm gene regulatory networks. Dev. Biol. 261:55–81.

Bailey, T. J., El-Hodiri, H., Zhang, L., Shah, R., Mathers, P. H., and Jamrich, M. (2004). Regulation of vertebrate eye development by Rx genes. Int. J. Dev. Biol. 48:761–770.

Brown, N. L., Kanckar, S., Vetter, M. L., Tucker, P. K., Gemza, D. L., and Glaser, T. (1998). Math5 encodes a murine basic helix-loop-helix transcription factor expressed during early stages of retinal neurogenesis. Development 125:4821–4833.

Brown, N. L., Patel, S., Brzezinski, J., and Glaser, T. (2001). Math5 is required for retinal ganglion cell and optic nerve formation. Development 128:2497–2508.

Burmeister, M., Novak, J., Liang, M. Y., Basu, S., Ploder, L., Hawes, N. L., Vidgen, D., Hoover, F., Goldman, D., et al. (1996). Ocular retardation mouse caused by Chx10 homeobox null allele: Impaired retinal progenitor proliferation and bipolar differentiation. Nat. Genet. 12:376–384.

Cayouette, M., Poggi, L., and Harris, W. A. (2006). Lineage in the vertebrate retina. Trends Neurosci. 29:563–570.

Coombs, J., van der List, D., Wang, G. Y., and Chalupa, L. M. (2006). Morphological properties of mouse retinal ganglion cells.

Neuroscience 140:123–136.

Dakubo, G. D., and Wallace, V. A. (2004). Hedgehogs and retinal ganglion cells: Organizers of the mammalian retina. Neuroreport 15:479–482.

Davidson, E. H., McClay, D. R., and Hood, L. (2003). Regulatory gene networks and the properties of the developmental process. Proc. Natl. Acad. Sci. U.S.A. 100:1475–1480.

Davidson, E. H., Rast, J. P., Oliveri, P., Ransick, A., Calestani, C., Yuh, C. H., Minokawa, T., Amore, G., Hinman, V., et al. (2002). A genomic regulatory network for development. Science 295:1669–1678.

Diao, L., Deng, Q., and He, S. (2004). Development of the mouse retina: Emerging morphological diversity of the ganglion cells.

J. Neurobiol. 61:236–249.

Dyer, M. A. (2003). Regulation of proliferation, cell fate specification and differentiation by the homeodomain proteins Prox1, Six3, and Chx10 in the developing retina. Cell Cycle 2: 350–357.

Erkman, L., McEvilly, R. J., Luo, L., Ryan, A. K., Hooshmand, F., O’Connell, S. M., Keithley, E. M., Rapaport, D. H., Ryan, A. F., et al. (1996). Role of transcription factors Brn-3.1 and Brn-3.2 in auditory and visual system development. Nature 381:603–606.

Erkman, L., Yates, P. A., McLaughlin, T., McEvilly, R. J., Whisenhunt, T., O’Connell, S. M., Krones, A. I., Kirby, M. A., Rapaport, D. H., et al. (2000). A POU domain transcription factor-dependent program regulates axon pathfinding in the vertebrate visual system. Neuron 28:779–792.

Feng, L., Xie, X., Joshi, P. S., Yang, Z., Shibasaki, K., Chow, R. L., and Gan, L. (2006). Requirement for Bhlhb5 in the specification of amacrine and cone bipolar subtypes in mouse retina. Development 133:4815–4825.

Fu, X., Kiyama, T., Mu, X., and Klein, W. (2008). [Shh is a target gene for Pontf2 in developing retinal ganglion cells.] Unpublished results.

Fu X., Sun, H., Klein, W. H., and Mu, X. (2006). Beta-catenin is essential for lamination but not neurogenesis in mouse retinal development. Dev. Biol. 299:424–437.

Fujitani, Y., Fujitani, S., Luo, H., Qiu, F., Burlison, J., Long, Q., Kawaguchi, Y., Edlund, H., Macdonald, R. J., et al. (2006). Ptf1a determines horizontal and amacrine cell fates during mouse retinal development. Development 133:4439–4450.

Furukawa, T., Morrow, E. M., and Cepko, C. L. (1997). Crx, a novel otx-like homeobox gene, shows photoreceptor-specific expression and regulates photoreceptor differentiation. Cell 91: 531–541.

Furuta, Y., Lagutin, O., Hogan, B. L., and Oliver, G. C. (2000). Retinaand ventral forebrain-specific Cre recombinase activity in transgenic mice. Genesis 26:130–132.

Gan, L., Wang, S. W., Huang, Z., and Klein, W. H. (1999). POU domain factor Brn-3b is essential for retinal ganglion cell differentiation and survival but not for initial cell fate specification. Dev. Biol. 210:469–480.

Gan, L., Xiang, M., Zhou, L., Wagner, D. S., Klein, W. H., and Nathans, J. (1996). POU domain factor Brn-3b is required for the development of a large set of retinal ganglion cells. Proc. Natl. Acad. Sci. U.S.A. 93:3920–3925.

Harpavat, S., and Cepko, C. L. (2003). Thyroid hormone and retinal development: An emerging field. Thyroid 13:1013–1019.

Hashimoto, T., Zhang, X. M., Chen, B. Y., and Yang, X. J. (2006). VEGF activates divergent intracellular signaling components to regulate retinal progenitor cell proliferation and neuronal differentiation. Development 133:2201–2210.

Hatakeyama, J., and Kageyama, R. (2004). Retinal cell fate determination and bHLH factors. Semin. Cell Dev. Biol. 15:83–89.

Hattar, S., Liao, H. W., Takao, M., Berson, D. M., and Yau, K. W. (2003). Melanopsin-containing retinal ganglion cells: Architecture, projections, and intrinsic photosensitivity. Science 295:1065–1070.

Hutcheson, D. A., Hanson, M. I., Moore, K. B., Brown, N. L., and Vetter, M. L. (2005). bHLH-dependent and -independent modes of Ath5 gene regulation during retinal development. Development 132:829–839.

Hutcheson, D. A., and Vetter, M. L. (2001). The bHLH factors Xath5 and XNeuroD can upregulate the expression of XBrn3d, a POU-homeodomain transcription factor. Dev. Biol. 232:327– 338.

Ikeda, H., Osakada, F., Watanabe, K., Mizuseki, K., Haraguchi, T., Miyoshi, H., Kamiva, D., Honda, Y., et al. (2005). Generation of RX+/Pax6+ neural retinal precursors from embryonic stem cells. Proc. Natl. Acad. Sci. U.S.A. 102:11331– 11336.

Jadhav, A. P., Cho, S. H., and Cepko, C. L. (2006a). Notch activity permits retinal cells to progress through multiple progenitor states and acquire a stem cell property. Proc. Natl. Acad. Sci. U. S.A. 103:18998–19003.

Jadhav, A. P., Mason, H. A., and Cepko, C. L. (2006b). Notch1 inhibits photoreceptor production in the developing mammalian retina. Development 133:913–923.

Ji H., Vokes, S. A., and Wong, W. H. (2006). A. comparative analysis of genome-wide chromatin immunoprecipitation data for mammalian transcription factors. Nucleic Acids Res. 34:e146.

Kay, J. N., Finger-Baier, K. C., Roeser, T., Staub, W., and Baier, H. (2001). Retinal ganglion cell genesis requires lakritz, a zebrafish atonal homolog. Neuron 30:725–736.

Kay, J. N., Link, B. A., and Baier, H. (2005). Staggered cellintrinsic timing of ath5 expression underlies the wave of ganglion cell neurogenesis in the zebrafish retina. Development 132: 2573–2585.

Kim, J., Wu, H. H., Lander, A. D., Lyons, K. M., Matzuk, M. M., and Calof, A. L. (2005). GDF11 controls the timing of

progenitor cell competence in developing retina. Science 308: 1927–1930.

Lamba, D. A., Karl, M. O., Ware, C. B., and Reh, T. A. (2006). Efficient generation of retinal progenitor cells from human embryonic stem cells. Proc. Natl. Acad. Sci. U.S.A. 103:12769– 12774.

Le, T. T., Wroblewski, E., Patel, S., Riesenberg, A. N., and Brown, N. L. (2006). Math5 is required for both early retinal neuron differentiation and cell cycle progression. Dev. Biol. 295:764–778.

Levine, M., and Davidson, E. H. (2006). Gene regulatory networks for development. Proc. Natl. Acad. Sci. U.S.A. 102:4936–4942.

Li S., Mo Z., Yang, X., Price, S. M., Shen, M. M., and Xiang, M. (2004). Foxn4 controls the genesis of amacrine and horizontal cells by retinal progenitors. Neuron 43:795–760.

Liu, H., Thurig, S., Mohamed, O., Dufort, D., and Wallace, V. A. (2006). Mapping canonical Wnt signaling in the developing and adult retina. Invest. Ophthalmol. Vis. Sci. 47:5088– 5097.

Livesey, F. J., and Cepko, C. L. (2001). Vertebrate neural cell-fate determination: Lessons from the retina. Nat. Rev. Neurosci. 2:109–118.

Locker, M., Agathocleous, M., Amato, M. A., Parain, K., Harris, W. A., and Perron, M. (2006). Hedgehog signaling and the retina: Insights into the mechanisms controlling the proliferative properties of neural precursors. Genes Dev. 20:3036–3048.

Logan, M. A., Steele, M. R., Van Raay, T. J., and Vetter, M. L. (2005). Identification of shared transcriptional targets for the proneural factors Xath5 and XNeuroD. Dev. Biol. 285: 570–583.

MacLaren, R. E., Pearson, R. A., MacNeil, A., Douglas, R. H., Salt, T. E., Akimoto, M., Swaroop, A., Sowden, J. C., and Ali, R. R. (2006). Retinal repair by transplantation of photoreceptor precursors. Nature 444:203–207.

Mao, C. A., Kiyama, T., Pan, P., Furuta, Y., Hadjantonakis, A. K., and Klein, W. H. (2008). Eomesodermin, a target gene of Pou4f2, is required for retinal ganglion cell and optic nerve development in the mouse. Development 135:271–280.

Marquardt, T. (2003). Transcriptional control of neuronal diversification in the retina. Prog. Retin. Eye. Res. 22:567–577.

Marquardt, T., Ashery-Padan, R., Andrejewski, N., Scardigli, R., Guillemot, F., and Gruss, P. (2001). Pax6 is required for the multipotent state of retinal progenitor cells. Cell 105:43– 55.

Marquardt, M., and Gruss, P. (2002). Generating neuronal diversity in the retina: One for nearly all. Trends Neurosci 25:32–38.

Martin, S. E., Mu, X., and Klein, W. H. (2005). Identification of an N-terminal transcriptional activation domain within Brn3b/ POU4f2. Differentiation 73:18–27.

Masai, I., Yamaguchi, M., Tonou-Fujimori, N., Komori, A., and Okamoto, H. (2005). The hedgehog-PKA pathway regulates two distinct steps of the differentiation of retinal ganglion cells: The cell-cycle-exit of retinoblasts and their neuronal maturation. Development 132:1539–1553.

Matter-Sadzinski, L., Matter, J. M., Ong, M. T., Hernandez, J., and Ballivet, M. (2001). Specification of neurotransmitter receptor identity in developing retina: The chick ATH5 promoter integrates the positive and negative effects of several bHLH proteins. Development 128:217–231.

Matter-Sadzinski, L., Puzianowska-Kuznicka, M., Hernandez, J., Ballivet, M., and Matter, J. M. (2005). A. bHLH transcriptional network regulating the specification of retinal ganglion cells. Development 132:1907–1921.

McLaughlin, T., Hindges, R., and O’Leary, D. D. (2003). Regulation of axial patterning of the retina and topographic mapping to the brain. Curr. Opin. Neurobiol. 13:57–69.

Mears, A. J., Kundo, M., Swain, P. K., Takada, Y., Bush, R. A., Saunders, T. L., Sieving, P. A., and Swaroop, A. (2001). Nrl is required for rod photoreceptor development. Nat. Genet. 29: 447–453.

Mu, X., Beremand, P. D., Zhao, S., Pershad, R., Sun, H., Scarpa, A., Liang, S., Thomas, T. L., and Klein, W. H. (2004). Discrete gene sets depend on POU domain transcription factor Brn3b/ Brn-3.2/POU4f2 for their expression in the mouse embryonic retina. Development 131:1197–1210.

Mu, X., Fu, X., Beremand, P. D., Thomas, T. L., and Klein, W. H. (2008). Gene regulation logic in retinal ganglion cell development: Isl1 defines a critical branch distinct from but overlapping with Pou4f2. Submitted for publication.

Mu, X., Fu, X., Sun, H., Beremand, P. D., Thomas, T. L., and Klein, W. H. (2005a). A. gene network downstream of transcription factor Math5 regulates retinal progenitor cell competence and ganglion cell fate. Dev. Biol. 280:467–481.

Mu, X., Fu, X., Sun, H., Liang, S., Maeda, H., Frishman, L. J., and Klein, W. H. (2005b). Ganglion cells are required for normal progenitor-cell proliferation but not cell-fate determination or patterning in the developing mouse retina. Curr. Biol. 15: 525–530.

Mu, X., and Klein, W. H. (2004). A. gene regulatory hierarchy for retinal ganglion cell specification and differentiation. Semin. Cell Dev. Biol. 15:115–123.

Mumm, J. S., Williams, P. R., Godinho, L., Koerber, A., Pittman, A. J., Roeser, T., Chien, C. H., Baier, H., and Wong, R. O. (2006). In vivo imaging reveals dendritic targeting of laminated afferents by zebrafish retinal ganglion cells. Neuron 52:609– 621.

Nelson, B. R., Gumusen, B., Hartman, B. H., and Reh, T. A. (2006). Notch activity is downregulated just prior to retinal ganglion cell differentiation. Dev. Neurosci. 28:128–141.

Neumann, C. J. (2005). Hedgehogs as negative regulators of the cell cycle. Cell Cycle 4:1139–1140.

Neumann, C. J., and NÜesslein-Volhard, C. (2000). Patterning of the zebrafish retina by a wave of sonic hedgehog activity. Science 289:2137–2139.

Ohtoshi, A., Wang, S. W., Maeda, H., Saszik, S. M., Frishman, L. J., Klein, W. H., and Behringer, R. R. (2004). Regulation of retinal cone bipolar cell differentiation and photopic vision by the CVC homeobox gene Vsx1. Curr. Biol. 14:530– 536.

Ouchi, Y., Tabata, Y., Arai, K., and Watanabe, S. (2005). Negative regulation of retinal-neurite extension by beta-catenin signaling pathway. J. Cell Sci. 118:4473–4483.

Perron, M., and Harris, W. A. (2000). Determination of vertebrate retinal progenitor cell fate by the Notch pathway and basic helix-loop-helix transcription factors. Cell Mol. Life Sci. 57:215–223.

Poggi, L., Vitorino, M., Masai, I., and Harris, W. A. (2005). Influences on neural lineage and mode of division in the zebrafish retina in vivo. J. Cell Biol. 171:991–999.

Rodriguez, J., Esteve, P., Weini, C., Ruiz, J. M., Fermin, Y., Trousse, F., Dwivedy, A., Holt, C., and Bovolenta, P. (2005). SFRP1 regulates the growth of retinal ganglion cell axons through the Fz2 receptor. Nat. Neurosci. 8:1281–1282.

Schmitt, A. M., Shi, J., Wolf, A. M., Lu, C. C., King, L. A., and Zou, Y. (2006). Wnt-Ryk signaling mediates medial-lateral retinotectal topographic mapping. Nature 439:31–37.

Sicinski, P., Donaher, J. L., Parker, S. B., Li, T., Fazeli, A., Gardner, H., Haslam, S. Z., Bronson, R. T., Elledge, S. J., et al. (1995). Cyclin D1 provides a link between development and oncogenesis in the retina and breast. Cell 82:621–630.

Silver, S. J., and Rebay, I. (2005). Signaling circuitries in development: Insights from the retinal determination gene network.

Development 132:3–13.

Taranova, O. V., Magness, S. T., Fagan, B. M., Wu, Y., Surzenko, N., Hutton, S. R., and Pevny, L. H. (2006). SOX2 is a dose-dependent regulator of retinal progenitor competence. Genes Dev. 20:1187–1202.

Trieu, M., Rhee, J. M., Fedtsova, N., and Turner, E. E. (1999). Autoregulatory sequences are revealed by complex stability screening of the mouse Brn-3.0 locus. J. Neurosci. 19:6549– 6558.

Vetter, M. L., and Brown, N. L. (2001). The role of basic helix- loop-helix genes in vertebrate retinogenesis. Semin. Cell Dev. Biol. 12:491–498.

Wang, J. C., and Harris, W. A. (2005). The role of combinatorial coding by homeodomain and bHLH transcription factors in retinal cell fate specification. Dev. Biol. 285:101–115.

Wang, S. W., Gan, L., Martin, S. E., and Klein, W. H. (2000). Abnormal polarization and axon outgrowth in retinal ganglion cells lacking the POU-domain transcription factor Brn-3b. Mol. Cell. Neurosci. 16:141–156.

Wang, S. W., Kim, B. S., Ding, K., Wang, H., Sun, D., Johnson, R. L., Klein, W. H., and Gan, L. (2001). Requirement for math5 in the development of retinal ganglion cells. Genes Dev. 15:24–29.

Wang, S. W., Mu, X., Bowers, W. J., Kim, D. S., Plas, D. J., Crair, M. C., Federoff, H. J., Gan, L., and Klein, W. H. (2003). Brn3b/Brn3c double knockout mice reveal an unsuspected role for Brn3c in retinal ganglion cell axon outgrowth. Development 129:467–477.

Wang, Y., Dakubo, G. D., Thurig, S., Mazerolle, C. J., and Wallace, V. A. (2005). Retinal ganglion cell–derived sonic hedgehog locally controls proliferation and timing of RGC development in the embryonic mouse retina. Development 132: 5103–5113.

Wee, R., Castrucci, A. M., Provencio, I., Gan, L., and Van Gelder, R. N. (2003). Loss of photic entrainment and altered free-running circadian rhythms on math5−/− mice. J. Neurosci. 22:10427–10433.

Yan, R. T., Ma, W., Liang, L., and Wang, S. Z. (2005). bHLH genes and retinal cell fate specification. Mol. Neurobiol. 32: 157–171.

Yang, Z., Ding, K., Pan, L., Deng, M., and Gan, L. (2003). Math5 determines the competence state of retinal ganglion cell progenitors. Dev. Biol. 264:240–254.

Yaron, O., Farhy, C., Marquardt, T., Applebury, M., and Ashery-Padan, R. (2006). Notch1 functions to suppress conephotoreceptor fate specification in the developing mouse retina. Development 133:1367–1378.

Zeller, K. I., Zhao, X., Lee, C. W., Chiu, K. P., Yao, F., Yustein, J. T., Ooi, H. S., Orlov, Y. L., Shahab, A., et al. (2006). Global mapping of c-Myc binding sites and target gene networks in human B. cells. Proc. Natl. Acad. Sci. U.S.A. 103:17834–17839.

Zhang, L., Mathers, P. H., and Jamrich, M. (2000). Function of Rx, but not Pax6, is essential for the formation of retinal progenitor cells in mice. Genesis 28:135–142.

Zhao, X., Liu, J., and Ahmad, I. (2002). Differentiation of embryonic stem cells into retinal neurons. Biochem. Biophys. Res. Commun. 297:177–184.

Cell loss is commonly observed in the development of vertebrate neural structures. In many regions of the developing nervous system, neurons and their connections are produced in excess of their adult number and then partly eliminated (reviewed in Oppenheim, 1981; Buss and Oppenheim, 2004). In the developing retina, cell death was initially analyzed in vertebrates other than the mouse because their retinal structures were more accessible to manipulation, but in recent years the use of transgenic technologies has made the mouse the elective model in which to test the role of individual molecular components in the death process. In this chapter we refer to both mouse and nonmouse studies to provide an overall view on cell death in the retina. The first part of the chapter describes the time course and locations of cell death in the developing mouse retina; the second part is a brief summary of key experiments investigating the contribution of cell death to circuit maturation. The third part summarizes the role of neurotrophic factors in controlling cell death in retinal development, and the last part looks at current knowledge on the genes and molecules involved in retinal cell death.

Major phases of cell death in the developing mouse retina

Dying cells can be recognized on the basis of morphological or biochemical criteria, the latter being a very powerful tool to identify specific death pathways, the former being most handy for preliminary or quantitative analyses. On light microscopy, neurons that die during normal development exhibit condensation and breakdown of chromatin into fragments, or apoptotic bodies (Wyllie et al., 1980). Cells displaying this morphology (figure 27.1A) are called pyknotic. Pyknotic cells have been observed throughout the course of development of the mouse retina (Silver and Hughes, 1973; Young, 1984; Laemle et al., 1999; Pequignot et al., 2003), with three major temporal peaks in terms of density of dying cells. An early phase of cell death affects the retina and the surrounding eye tissues between embryonic day 9 (E9) and E11, around the time of optic vesicle formation and retinal induction. A second peak of cell death is observed between E15.5 and E17.5, when neurogenesis is already under way. Finally, dying neurons are observed in different retinal locations during the first 2 weeks of life, when retinal neurons are differentiating and synaptic connections are being formed.

In the early phase of death, pyknotic cells in the retina are observed mostly in the central region and are likely to be proliferating cells, as almost no neuronal cell has been generated at this age. At this stage, cell death is also observed in the lens and in the non-neuronal epithelium behind the retina. It has been suggested that this early death might contribute to shape the eye, analogous to the role played by death during other morphogenetic events in the organism. In line with this view, both bone morphogenetic proteins (BMPs) and transforming growth factor-β (TGF-β), which have already been involved in morphogenetic events outside the nervous system, appear to regulate retinal cell death at this stage. In the chick, BMP signaling appears to regulate both cell death and cell proliferation around the time of optic vesicle formation (Trousse et al., 2001). In addition, mice deficient in the Msx2 gene, which is thought to regulate BMP signaling, have increased early death and develop microphthalmia as a consequence of the reduced pool of progenitors in the eye tissues (Wu et al., 2003). Finally, knockout mice for both TGF-β2 and TGF-β3 display a significant reduction in early cell death (Dunker and Krieglstein, 2002).

When the second peak of death occurs in the retina, dying cells appear to consist of both early-born retinal ganglion cells (RGCs) and neuroblasts. Pharmacological manipulations in the chick have shown that RGC death in this phase is triggered by nerve growth factor (NGF) binding to its lowaffinity receptor, p75 (Frade et al., 1996). In the mouse, many cells express p75 in and along the developing optic nerve at this stage, and death at this time is reduced in embryos carrying deletions in the NGF or the p75 gene (Frade and Barde, 1999). Furthermore, knockout mice for NRIF, a zinc finger protein that interacts with p75 signaling, have reduced cell death in the retina, as do NGF or p75 null mice (Casademunt et al., 1999). NGF/p75-mediated cell death has been found in several instances of neuronal development and in neurodegeneration (Dechant and Barde, 2002). In the developing nervous tissue, death triggered by NGF concentrates in regions that are subsequently crossed by growing nerve fibers (the central region of the ganglion cell layer [GCL] and the future optic nerve head in the retina), leading to the suggestion that this death might clear the way for newly forming axonal pathways (Frade et al., 1996; Frade and Barde, 1999). As for death in the neuroblast layer at this stage, its significance and amount are still largely unclear (reviewed in de la Rosa and de Pablo, 2000).

During the first 2 weeks of life in the mouse retina, dying cells are observed at different retinal locations (Young, 1984; Pequignot et al., 2003): at postnatal day 2 (P2), most dying cells are in the GCL; at P9, dying cells are observed mostly in the inner nuclear layer (INL); finally, at P15, dying cells are both in the GCL and in the outer nuclear layer (ONL). The observation of degenerating cells in various retinal layers during retinal differentiation is common to mammals and suggests that most cell populations in the retina undergo phases of cell loss while differentiating. However, verifying this supposition is proving difficult: degenerating cells can be identified on the basis of morphological or biochemical criteria, which commonly include nuclear alterations (see figure 27.1A), but these features last for only a brief time (ca. 1 hour in the retina; Cellerino et al., 2000), and furthermore, by the time nuclear morphology is altered, cells have usually lost the differentiation markers necessary to identify their original cell type. To bypass this problem, the total number of cells in some populations has been evaluated at different developmental ages and the amount of cell loss obtained as the difference between the peak and the final number. This may underestimate death contribution in cases where new cell addition and cell loss partially overlap in time, as observed at least in some retinal populations (Galli-Resta and Ensini, 1996; Seecharan et al., 2003; Resta et al., 2005), but it provides at least conservative estimates.

In the study of retinal cell death, much attention has been focused on RGCs. This privileged status is at least partly the result of more reliable cell counts for RGCs than for other developing neuronal populations (since RGCs project outside the retina, they can be retrogradely labeled, and often retain these labels when dying; in addition, RGC counts can be

validated by axonal counts in the optic nerve). As well, the accessibility of the retina to external manipulation has made RGCs a popular model not only for naturally occurring death in development, but also for neuronal death after axotomy or in different pathological conditions. On the basis of both RGC counts after injection of retrograde tracers in the RGC targets and axonal counts in the optic nerve, a significant decrement in the number of RGCs has been reported in developing mammalian retinas, ranging between 50% and 75% (figure 27.1B, reviewed in Dreher and Robinson, 1988). As we have already mentioned, this decrease in total cell number may underestimate the number of dying cells if death occurs while new cells are being added to the population. Attempts to clarify this matter have led to conflicting estimates of the number of dying RGCs (e.g., 50% for the earliest-born mouse RGCs: Farah and Easter, 2005; 69% for all mouse RGCs: Strom and Williams, 1998; up to 90% for rat GCL neurons: Galli-Resta and Ensini, 1996).

Contribution of cell death to the shaping of retinofugal and retinal circuits

The simultaneous occurrence of cell death and the formation of synaptic connections in the retina and elsewhere has led to the hypothesis that death might contribute to circuit refinements. In particular, death has been hypothesized as an elective process to (1) match the size of interconnected neuronal populations, (2) refine connections and remove projecting errors, and (3) control cell density. Here we consider studies addressing these questions in the development of retinal projections to the brain and in the maturation of intraretinal circuits.

Numerous investigations outside the visual system suggest that death contributes to matching the sizes of interconnected neuronal populations. In the visual system, there is strong experimental evidence that target cells are necessary for the survival of presynaptic cells during development, but these interactions do not lead to a precise quantitative matching between afferent and target cell populations at maturity. In adult monkeys, interindividual variations in the number of RGCs do not correlate with changes in the number of cells in the retinorecipient nuclei (Spear et al., 1996; Suner and Rakic, 1996). Similarly, analysis across adult specimens from 56 mouse strains showed no detectable correlation between the number of neurons in the populations of lateral geniculate neurons and RGCs (Seecharan et al., 2003). In contrast, RGC number correlates significantly with lateral geniculate glial cell number, but how cell death might contribute to this correlation remains to be explored.

The refinement of projection errors during development is a second process to which retinal cell death is hypothesized to contribute. In mammals, initial projections of RGCs to the lateral geniculate nucleus and the superior colliculus display some degree of imprecision. In particular, more cells project ipsilaterally or form topographically incorrect connections than in the adult (reviewed in Cowan et al., 1984). Blocking RGCs’ electrical activity during this period reduces the loss of wrongly projecting cells in various species (figure 27.2), indicating a role for cell death in this process (O’Leary et al., 1986b; Pequignot and Clarke, 1992). However, the amount and time course of RGC loss are not affected (O’Leary et al., 1986a), or only marginally so (Scheetz et al., 1995), suggesting that normally, retinal impulse activity makes erroneously projecting RGCs more likely to die, but does not otherwise affect the death process.

It has also been suggested that cell death might contribute to shaping the intraretinal circuitry. RGC removal following optic nerve transection at birth in rats did not significantly affect the density of pyknotic profiles in the remaining retinal layers (Beazley et al., 1987), but when specific amacrine cell populations were analyzed in the adult ferret retina after neonatal optic nerve section, these appeared to be differentially affected, some being increased and others decreased in the absence of RGCs. Thus, cell-cell interactions controlling survival in retinal development appear highly specific for the various cell types, and death might indeed contribute to the fine regulation of cell number in the different populations (Williams et al., 2001).

A particular role of cell death during retinal maturation may indeed be in the control of cell density within individual neuronal populations. In the retina, most neuronal types are organized in orderly arrays known as mosaics, and this organization is thought to ensure even detection and processing of the images impinging on the retina (reviewed in Cook and

Chalupa, 2000; Galli-Resta, 2002). Schematically, mosaics can be viewed as cell arrays where density is finely regulated on a local scale, so that each neuron ends up being regularly spaced from its like neighbors. Theoretical studies indicate that death cannot by itself generate regular arrays (Eglen and Willshaw, 2002), but several investigations have shown that cell death contributes to the fine control of local cell density within specific retinal populations. First, it has been demonstrated that refinement of the ON alpha RGCs mosaic in cats coincides with the removal of 20% of these cells, and that blocking electrical activity contrasts the selectivity of cell elimination and leads to less orderly mosaics (Jeyarasasingam et al., 1998). Second, in transgenic mice in which retinal cell death is partially suppressed by Bcl-2 overexpression, the population of dopaminergic amacrine cells undergoes an almost 10-fold increase in cell density and a decrement in the regularity of cell spacing (Raven et al., 2003). Finally, when extracellular ATP signaling is prevented in the rat retina, death among the cholinergic amacrine cells is reduced, causing an increase in their density (figure 27.2B, Resta et al., 2005). In summary, mosaic formation appears to involve two types of processes: lateral cell movements, and

Figure 27.2 A, Cell death involvement in the refinement of topographic projections. On P0, both nasal (N) and temporal (T) RGCs project to the posterior contralateral colliculus in rats (left). The T contingent is about 14% of the N. The T contingent is almost completely eliminated by P12, after the period of cell death (center). However, if RGC impulse activity is blocked, the T contingent is maintained at the P0 value (right). (Adapted from O’Leary et al., 1986b.) B, Among the cholinergic amacrine cells, death induced by endogenous extracellular ATP contributes to reducing local cell density during development. Left, Normal cholinergic cell density. Right, Cholinergic cell density 24 hours after extracellular ATP degradation. (From Resta et al., 2005.)

possibly a spatial control on the location of cell genesis aimed at positioning the cell appropriately, while at the same time death eliminates cells, thus keeping local density under control.

The neurotrophic hypothesis

The study of cell death in the different neural regions of the developing organism, and in particular the finding that the survival of postmitotic neurons depends on intercellular interactions, classically exemplified by the need for a target tissue, and the identification of intrinsic trophic factors promoting neuronal survival (initiated with the discovery of NGF), has led to a general hypothesis about cell death regulation. This classic hypothesis in developmental neurobiology, often referred to as the neurotrophic hypothesis, states that during development and differentiation, neuronal survival is contingent on the supply of trophic factors provided by the environment, with a prominent role played by postand presynaptic tissues (reviewed in Oppenheim, 1991; Bennet et al., 2002; Davies, 2003).

In the retina, current evidence suggests that different neurotrophic factors may be accessible to particular neuronal populations in specific spatiotemporal sequences, so that several neurotrophic interactions may be required for normal development of each neuronal type (Korsching, 1993). Further, as we have already seen, there is also evidence of a death-promoting effect by a classic neurotrophin (endogenous NGF) during embryonic retinal development.

Neurotrophins and their receptors are expressed in the developing retina during the second and the late phases of naturally occurring cell death (reviewed in Frade et al., 1999). Most experimental manipulations concerning the trophic action of neurotrophins on cell death in the retina have been done in the chick. Exogenously applied brainderived neurotrophic factor (BDNF) reduces RGC death and increases the number of axons in the optic nerve (Frade et al., 1997). Functionally blocking antibodies against neu- rotrophin-3 (NT-3) prevents the death of a subset of amacrine and ganglion cells in cell culture from E9–E11 dissociated chick retinas (de la Rosa et al., 1994). In vivo, when endogenous NT-3 is neutralized, the retina is reduced in size and abnormal in its organization, showing a narrowing of the inner plexiform layer, most likely as a result of the reduced neuronal number in the adjacent layers (Bovolenta et al., 1996). In the rat, injecting BDNF or NT-4/NT-5 either into the superior colliculus or into the eye limits the death of neonatal RGCs (Cui and Harvey, 1994, 1995; Ma et al., 1998). Similarly, BDNF application has been shown to prevent rat RGC death in culture (reviewed in Frade et al., 1999). Other studies in culture, however, suggest that promoting RGC survival requires simultaneous stimulation by multiple trophic factors (Meyer-Franke et al., 1995). In

line with this hypothesis, null mutations for BDNF or NT-4, individually or together, and for their functional receptor TrkB do not display gross defects in the neural retina, suggesting a compensation by other trophic factors (reviewed in Snider, 1994). Finally, recent studies indicate that manipulating the levels of BDNF or its receptor TrkB can modulate the time course of death among RGCs (Pollock et al., 2003), but not their final number (Cellerino et al., 1997; Rohrer et al., 2001).

An additional factor linked to death modulations in the developing retina is insulin-like growth factor (IGF), which, together with IGF-I receptors, is expressed during retinal development (de Pablo and de la Rosa, 1995). Moreover, mRNA and protein expression of the IGF system exhibit temporal and spatial correlation with patterns of cell death during retinal development (Kleffens et al., 1999), and recent evidence suggests that IGF-I modulates RGC death in vivo (Gutierrez-Ospina et al., 2002). Interestingly, a recent study shows that environmental enrichment can modulate the temporal pattern of cell death in the GCL, most likely through modulation of endogenous IGF-I levels (Sale et al., 2007). In addition, it is well established that insulin receptor substrate 2 (Irs2) appears essential for photoreceptor survival in the mouse retina: compared to control littermates, Irs2 knockout mice lose 10% of their photoreceptors 1 week after birth and up to 50% by 2 weeks of age as a result of increased apoptosis (Yi et al., 2005).

Both BMP and TGF-β appear to affect death not only in the earliest phase, but also during postnatal development. Mice deficient in the BMP receptor BMPrIb display higher than normal cell death rates in the INL on P7 (Liu et al., 2003). The addition of neutralizing anti-TGF-β antibodies reduced RGC death in cultured mouse retinas, corresponding to the second and the late death phases (Beier et al., 2006). Finally, several other factors, including interleukin-2 and -4 and TNF-α, have been shown to have neuroprotective effects in the developing retina, but their role in normal cell death, if any, is unclear (reviewed in Linden and Reese, 2006).

In summary, retinal cell survival cannot be linked to individual trophic factors. Individual factors appear to be more pleiotropic and promiscuous than once envisioned, and many interactions occur between members of different families of growth factors. It seems likely that a complex “growth factor homeostasis” occurs in the retina, with a network of factors regulating the survival and maintenance of retinal cells.

Molecules involved in signaling, regulating, and executing programmed cell death

The term programmed cell death was initially used to refer to death occurring normally during development, because of

its predictable time course and amount. This term has now acquired a more specific meaning, because the unveiling of molecular mechanisms underlying developmental death has made it clear that naturally occurring cell death in development depends on specific gene programs activated by the cells. Almost two decades of study in model organisms, neuronal cultures, and transgenic mice have revealed a central core of conserved molecules that activate, control, and execute the cell death program in many instances (Putcha and Johnson, 2004). In this “central dogma” of apoptosis (as death by activation of a genetic program is often but not always referred to), “thanatins” (Egl-1 in C. elegans, BH3-only proteins of the Bcl-2 family in vertebrates) are induced in cells destined to die. They interact with a “death inhibitor” (CED-9 in C. elegans, Bcl-2, Bcl-xL in vertebrates), thereby displacing an “adapter” (CED-4 in C. elegans, Apaf-1 in vertebrates), which can then promote activation of the “executioner” (CED-3 in C. elegans, caspases in vertebrates), which cleaves selected substrates and leads to cell death. This pathway appears remarkably conserved in evolution, but with several fundamental differences: (1) in vertebrates, caspase activation usually requires cytochrome-c release from mitochondria; (2) in multiple neuronal populations, cytochrome-c release requires expression of at least one multidomain proapoptotic Bcl-2 protein (e.g., Bax); and (3) finally, at least in some neuronal populations, caspase activation requires not only Bax-dependent cytochrome-c release but also inactivation of an inhibitor of apoptosis. It is important to point out that, as research proceeds, in addition to this highly conserved death program, alternative death pathways are also becoming known that appear to depend on activation of mechanisms other than the classic caspasedependent Bcl-2 family–controlled death program.

Evidence for the occurrence of the core apoptotic pathway in retinal cell death during development is still fragmentary, but a number of steps in the core pathways have correlates in the retina.

First, it has been established that naturally occurring cell death in the retina requires gene expression and protein synthesis, since intraocular injections of inhibitors of either transcription or synthesis block RGC death in the neonatal rat retina (Rabacchi et al., 1994).

Second, there is strong evidence for the involvement of both proand antiapoptotic proteins of the Bcl-2 family in cell death in the developing retina, although not all these studies support the classic view of the core apoptotic process. Transgenic mice carrying a deletion of the pro-apoptotic Bax gene show an almost doubled number of RGCs in adulthood (Mosinger Ogilvie et al., 1998). This correlates with a significant decrease in the percentage of apoptotic cells in the GCL at P2 (Pequignot et al., 2003). However, the overexpression of Bax driven by the neuron-specific enolase (NSE) promoter did not affect the numbers of RGCs (Bernard

et al., 1998), possibly for the late pattern of Bax expression it induced. Bax knockout mice also display lower rates of apoptosis in the INL between P8 and P11, and in the ONL on P15. Furthermore, either Bax or Bak deletion is sufficient for ectopic photoreceptor PCD during normal development (Hahn et al., 2003). Thus, the Bax-mediated pathway appears to play an important role in retinal cell death during development.

The picture concerning the antiapoptotic Bcl-2 members is more controversial. Gene deletion of the antiapoptotic Bcl-x in the mouse causes massive cell loss in the nervous system, but the retinal phenotype has not been reported. This knockout is lethal to the embryo. On the contrary, one transgenic line lacking the antiapoptotic Bcl-2 gene showed a 30% loss of RGCs, which, however, occurred after the physiological period of naturally occurring cell death (Cellerino et al., 1999). Naturally occurring RGC death is reduced in transgenic mice overexpressing Bcl-2 under the control of the NSE promoter (figure 27.3A–C; Bonfanti

Figure 27.3 A, Ventral view of the brain of a wild-type (left) and Bcl-2-overexpressing (right) adult mouse. The latter brain is larger than the wild-type mouse brain, and the optic nerve (ON) is clearly larger than normal, as it contains 60% more RGC axons than normal. This increase in RGC number is due to reduced death among these cells during postnatal development. B, A field in the wild-type P0 retina showing pyknotic cells (arrows). C, Almost no dying cell is observed at this stage in the Bcl-2-overexpressing retina. (B and C from Bonfanti et al., 1996.)